Research map / Image generation / Fast diffusion sampling
SalDiff-DTM: A Novel Dual-Temporal Modulated Diffusion Model for Omnidirectional Images Scanpath Prediction
Xiaohui Kong, Qian Liu, Dandan Zhu, Kaiwei Zhang, Xiongkuo Min
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add dual (semantic + image) GCN features (ablation) | SalDiff-DTM · Salient360! | REC: 4.334 -> 4.384 | Improved |
| add dual GCN features (ablation) | SalDiff-DTM · AOI | REC: 4.216 -> 4.226 | Improved |
| use only semantic-level or only image-level GCN instead of none (ablation) | SalDiff-DTM · Salient360! and AOI | REC: Salient360! 4.334 -> 4.317 / 4.331; AOI 4.216 -> 4.188 / 4.183 | Hurt |
| add BiLSTM feature (ablation) | SalDiff-DTM · Salient360! | REC: 4.340 -> 4.384 | Improved |
| add BiLSTM feature (ablation) | SalDiff-DTM · AOI | REC: 4.203 -> 4.226 | Improved |
| add both BiMamba and M3 blocks (TABiMamba) instead of neither (ablation) | SalDiff-DTM · Salient360! | REC: 4.331 -> 4.384 | Improved |
| add both BiMamba and M3 blocks instead of neither (ablation) | SalDiff-DTM · AOI | REC: 4.177 -> 4.226 | Improved |
| use SalDiff-DTM instead of ScanDMM | scanpath predictor · Salient360! | REC: 3.403 -> 4.384 | Improved |
| use SalDiff-DTM instead of ScanDMM | scanpath predictor · Sitzmann | DTW: 1953.319 -> 1984.446 | Hurt |
| use SalDiff-DTM instead of Pathformer3D | scanpath predictor · Salient360! | DTW: 1645.650 -> 1552.055 | Improved |
| use SalDiff-DTM instead of ScanDMM (efficiency) | scanpath predictor · Sitzmann test | inference latency (s): 0.2887 -> 0.2537 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ScanDMM — ScanDMM: A Deep Markov Model of Scanpath Prediction for 360deg Images (CVPR 2023) · better on 8 of 9 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- DiffEye: Diffusion-Based Continuous Eye-Tracking Data Generation Conditioned on Natural Images (NeurIPS 2025 · similarity 0.88)
- ST-DiffEye: Diffusion-based Continuous Gaze Generation via Joint Scanpath-Trajectory Modeling (NeurIPS 2026 · similarity 0.86)
- Diff-2-in-1: Bridging Generation and Dense Perception with Diffusion Models (ICLR 2025 · similarity 0.86)
- DiffBEV: Conditional Diffusion Model for Bird’s Eye View Perception (AAAI 2024 · similarity 0.86)
- MultiDiff: Consistent Novel View Synthesis from a Single Image (CVPR 2024 · similarity 0.86)
- StateSpaceDiffuser: Bringing Long Context to Diffusion World Models (NeurIPS 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
